The Microsoft AI research division unintentionally published 38TB of critical information while posting a container of open-source training data on GitHub, according to cybersecurity company Wiz.
Secrets, private keys, passwords, and more than 30,000 internal Microsoft Teams communications were discovered in a disk backup of the workstations of two workers that was made public by the disclosed data.
Wiz emphasized that because Microsoft does not offer a centralized method to manage SAS tokens within the Azure interface, it is difficult to track them.
Microsoft claimed that the data lead did not reveal customer data, that no customer data was leaked, and that this vulnerability did not put any internal services at risk.
Source: https://securityaffairs.com/151004/data-breach/microsoft-ai-data-leak.html
"id": "MIC33924923",
"linkid": "microsoft-ai",
"type": "Data Leak",
"date": "09/2023",
"severity": "60",
"impact": "3",
"explanation": "Attack with significant impact with internal employee data leaks"